predict后报错could not compute empirical Bayes means

gsem(Support-> c_housekeepingintensity  c_lifecare c_contactchildren c_contactfrequency,ologit) (Frailty->c_b1 c_b2 c_b3,ologit ) (Support internetuse -> Frailty), noestimate
matrix b = e(b)
matrix list b
matrix b[1,10] = 500
gsem(Support-> c_housekeepingintensity  c_lifecare c_contactchildren c_contactfrequency,ologit) (Frailty->c_b1 c_b2 c_b3,ologit ) (Support internetuse ->Frailty), noestimate from(b)
predict support,latent(Support)
(option ebmeans assumed)
(using 7 quadrature points)
could not compute empirical Bayes means;
missing values were returned by the evaluator
r(322);


结构方程初始值不行，无法运行，就使用了修改初始值的命令，后面想通过predict把潜变量Support化潜为显，但是报错322，could not compute empirical Bayes means; missing values were returned by the evaluator

希望有知道的朋友可以帮忙解答，感激！